Time-frequency plot produced when the Dynamic Auto-Regression (DAR) routine in CAPTAIN is applied to a ‘chirp’ signal. The analysis is fully automatic and based on the Kalman Filter and Fixed Interval Smoothing algorithms, with the ‘hyper para-meters’ optimized by maximum likelihood using prediction error decomposition.

This Website is intended primarily to support the CAPTAIN Toolbox for Matlab, which provides access to novel, mainly recursive,  algorithms for various important  aspects of identification, estimation, nonstationary time series analysis and signal processing, adaptive forecasting and automatic control system design. These have been developed between 1981 and the present at Lancaster University, UK, based on research carried out by Peter Young with colleagues at Lancaster, the Australian National University, Canberra, Australia (1975-1981) and the University of Cambridge, UK (1965-1975).

CAPTAIN’s numerous Matlab routines (m-files) provide access to most of the algorithms that are used in the Data-Based Mechanistic (DBM) approach to the modelling of stochastic, dynamic systems from sampled data. An overview of CAPTAIN can be obtained from the download page. A fully functional version of the CAPTAIN Toolbox can be downloaded from:

An optional e-Handbook is available but there is a charge for this..

The present Website also provides information that may be of use to CAPTAIN users: first, short articles on Technical Matters that discuss new and proposed developments in the Toolbox, as well as advice on the use and application of its routines; second Thoughts and Speculations, presenting my views on various topics that I am considering at the time.

Peter Young, Lancaster 22 June 2012

The Computer-Aided Program for Time-series Analysis and Identification of Noisy Systems (CAPTAIN) Toolbox

Website constructed by PCY using iWeb